The Agent Skills Directory

[COMMAND_EXECUTION]: The skill defines several shell commands to facilitate the workflow, including directory listing (ls docs/01_BRD/ ...) and the execution of local Python and Bash scripts (python ai_dev_ssd_flow/10_TSPEC/scripts/validate_utest.py, bash ai_dev_ssd_flow/10_TSPEC/scripts/validate_all_tspec.sh). These operations are restricted to paths within the repository's workspace and are consistent with its purpose as a development tool.\n- [DATA_EXPOSURE]: The skill accesses local documentation files (Layers 1-9) to generate Layer 10 artifacts. It does not perform any network operations, hardcode credentials, or access sensitive system directories (e.g., SSH keys or environment secrets).\n- [PROMPT_INJECTION]: The skill uses directive language such as 'CRITICAL' and 'MANDATORY' to guide the agent through the artifact creation process. These markers are used for workflow enforcement and do not attempt to bypass safety filters, extract system prompts, or override the agent's core instructions.\n- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it ingests content from external markdown files (upstream artifacts) and uses it to generate new documentation. While it has the capability to modify files via the '--auto-fix' flag in its validation scripts, these scripts are local to the repository, and the behavior is typical for automated documentation assistants.

doc-tspec